Figure 5-13. The 5.5 Å binding site of Sulpiride to the human D4 dopamine receptor. The numbers in the brackets indicate the TM helices to which the residues belong to.
Comparisons of Binding Site of Agonists vs. Antagonists: Experimental studies have
outlined the binding site for agonists and antagonists. The putative agonist-binding site is located between TM3, 4, 5, & 6, and some residues in the EC2 loop (assuming this loop closes during the process of activation). We have classified antagonists into two categories: Class I & II. Class I antagonists, such as Clozapine bind in the putative agonist binding pocket meaning they occupy the void between TM3, 4, 5 and 6. Class II antagonists, such as Haloperidol consists of two aromatic domains connected by a linker, which possess a protonated amino group. These antagonists bind in the cavity between
TM2, 3, 4, 5, 6, and 7. There are two aromatic micro domains (pictures of aromatic micro domains goes here). The first aromatic micro-domain is located in TM4 and TM6, which is composed of the very symmetric Trp160 (4), Ala164 (mutated in this receptor) (4), Trp407 (6), and Phe411 (6). This aromatic micro-domain stabilizes one of the aromatic rings of the class II antagonists. TM3 provides the salt bridge to TM3 Asp115, which stabilizes the ligand in place. TM5 provides weak interactions with heteroatom functionalities such as halogens on the rings of class II antagonists. These interactions are weak but important in recognition of the correct aromatic domain for docking into the cavity. The second aromatic micro-domain composed of Phe91 (2), Leu111 (mutated in this receptor) (3), Trp435 (7), and Tyr438 (7) stabilize the second aromatic ring group of the class II antagonists.
Comparison of agonists binding site: There is very little difference in the putative agonist-binding site of different agonists. The agonists studied, (dopamine, 7-OH DPAT, Apomorphine) all possess protonated amino groups, which are salt bridged to TM3 Asp115. All of these ligands studied form favorable hydrogen bonding interactions to a network of TM5 serines, although this is not an absolute necessary for agonism. There is, in every case studied, favorable interaction from the first aromatic micro-domain located in TM4 and TM6. A residue that has not been appreciated in drug design is His414 (6).
The presence of this residue contributes greatly to ligand binding, yet it appears that there few agonists utilize this residue effectively. It is important to note that the agonists studied effectively interact with both Ser196 and Ser200 in TM5. The interactions to Ser197, the third TM5 serine, are too long to be considered a hydrogen bond (on the
order of ~ 5 Å). Under no circumstance, due to structural constraints, can all three serines effectively interact with the agonists studied. It appears that there could at most be two hydrogen bonds to the TM5 serines. Although in our structure Ser197 is not participating in any interactions, it is possible that in a slightly different structure, perhaps one resulting from activation, there could be interactions to Ser197 and Ser200 as opposed to Ser196 and Ser200. All agonists studied cause strong coupling of TM3 and 5. None of the agonists studied block TM3 and 6 motions. Based on structural studies of rhodopsin, it has been established that a motion between TM3 and 6 are essential for activation. The coupling of TM3 and 6 by agonists causes a decrease in distance between TM3 and 5 while allowing for motion between TM3 and 6.
Comparison of antagonists binding site: The antagonists studied were classified into
two categories: Class I antagonists (exemplified by Clozapine), which bind in the putative agonist binding pocket; and class II antagonists (exemplified by Haloperidol), which bind in the cavity between TM2, 3, 4, 5, 6, and 7.
Class I Antagonists: the Clozapine-like antagonists salt bridge to the TM3 Asp115 with their protonated amino group. They do not have two aromatic rings and therefore only utilize the first aromatic micro-domain between TM4 and 6. As is the case with both Class I and II antagonists, they form only one weak interaction with the TM6 serine network. In our models the interaction may be with Ser196 or Ser200. At first glance, it appears that both the number of and strength of the interactions with the TM5 serines
may be important for activation. Interestingly, however, the situation is more complicated than it appears; Strange et al. have identified agonism from a non-hydroxylated form of the DPAT series. The critical distinguishing feature of an agonist vs. an antagonist appears to be its relative position to TM6. Class I antagonists burry their aliphatic domain deep into the conserved TM6 WXXFF motif. Experimental studies of rhodopsin suggest that motion of TM3 and 6 is necessary for activation. The binding of antagonists at the TM6 WXXFF would prevent any motion, specifically the hinge motion between TM3 and TM6 that is necessary for activation. It appears that the presence of one hydroxyl/one hydrogen bond donor/acceptor is not an absolute necessity for antagonism. The class I antagonists are further stabilized by Leu111 in TM3.
Class II Antagonists: the Haloperidol-like antagonists salt bridge to TM3 Asp115 with their protonated amino group. They possess two aromatic ring units. In most cases, only one of the aromatic rings is halogenated, and this is the ring that binds to the first aromatic micro-domain, with the second ring binding in the second aromatic micro- domain. The halogenated ring binds effectively into the cavity between TM4 and 6 and forms weak interactions with either Ser193 or Ser197 in TM5. It gains stability from the presence of TM4 Trp160 and Ala164 and TM6 Trp407 and Phe411. In some cases His414 (6) may also stabilize the class II antagonists. The non-halogenated aromatic domain is located in the void between TM2 and TM7. It gains stabilization from Phe91 (2), Leu111 (3), Trp435 (7), and Tyr438 (7). As is the case with the class I antagonists, class II antagonists prevent motion between TM3 and TM6 by burying their aliphatic portion between TM3 and TM6, thereby preventing interaction of these helices. There is
very little difference between the binding sites of the class II antagonists, although some utilize the cavity better than others.
Table 5-2. Binding energies in kcal/mol for a library of 10 ligands to the human D4
dopamine receptor.
D4 Receptor Experimental Ranking:
Spiperone~Haldol~Apomorphine>Clozapine~Sulpiride~Dopamine~Fenoldopam>
Raclopride~SCH23390~7-OHDPAT~SKF38393
D4 Receptor Theory Ranking:
Haldol~Spiperone~Apomorphine>Sulpiride~SCH23390~Dopamine>Fenoldopam~
SKF38393~Clozapine~Raclopride~7-OHDPAT
Ligand B.E. (Kcal/mol)
7OHDPAT -38.8
Apomorphine -50.1 Clozapine -33.1
Dopamine -47
Fenoldopam -33.9
Haldol -65.5
Raclopride -39.5
SCH -46.6
SKF -35.7
Spiperone -56.5 Sulpiride -49.8